The official Ruby Calibration Q/A Thread - Page 4 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #91 of 615 Old 01-25-2006, 06:26 AM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
As mentioned previously I am unable to get valid readings for my primaries so I cannot dial them in (or even attempt to) to the rec 709 coords.

Many of you have stated that at least as a first step in adjusting the primaries, you had to reduce the color control. Can you please let me know what you moved it down to from the default setting of 50?

For those of us that cannot dial in rec 709 using instrumentation, would we be at least somewhat more accurate by reducing the control control by this amount?

Currently I am using the Normal colorspace with Color/Tint at 50. I am certain that I have a D65 calibration from 30-100 with 0-3 dE.

However flesh tones do not look entirely accurate to me. They are a bit rosey and perhaps a bit yellow at time. On the contrary, my D65 calibration on the Sharp 10K resulted in much more realistic skin tones.

Do you guys have any suggestions on how to improve flesh tones?

Greg - did you feel that flesh tones looked very accurate on your Ruby review unit?
lovingdvd is offline  
Sponsored Links
Advertisement
 
post #92 of 615 Old 01-25-2006, 06:54 AM
AVS Special Member
 
JeffY's Avatar
 
Join Date: Mar 2001
Location: England
Posts: 3,667
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Well Bill you are confused because you are talking about luma levels yet continue to reference it as colorspace. Just to be clear, 0 - 100 is IRE levels, 16 - 235 is for component luma levels and 16 - 240 is for component chroma levels and 0-255 is for RGB levels.

And yes you still have to calibrate and match the correct levels for each source, this is where a video processor comes in very handy. Not doing so is a complete waste of money for such a hgh end display.
JeffY is offline  
post #93 of 615 Old 01-25-2006, 06:58 AM
AVS Special Member
 
bblue's Avatar
 
Join Date: Nov 2001
Location: El Cajon, CA
Posts: 1,051
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by lovingdvd View Post

First question: In this scenario is ColorFacts using 0-255 or 16-235?

Colorfacts running on the desktop will always be using 0-255.

Quote:


And is the Ruby using 0-255 or 16-235?

It depends. You'd always want it to be in the 0-255 space, but it will only be there in DVI IF you are running 1080p, and IF you have the input mode set to Computer. In fact, once you have the HTPC running 1080p over DVI and the Ruby is set to it, you can freely switch colorspace by changing the input mode between Video GBR (16-235) and Computer (0-255).

Quote:


Now the calibration is over, so I hook the Ruby back up to my Bravo D1 outputting 720p. Further assume just for discussion purposes that the Bravo is perfectly accurately and doesn't introduce any errors into the equation.

In this above scenario, is my GRAYSCALE part of the calibration still applicable for the Bravo? Or does the fact that I calibrated grayscale with 0-255 and now I'm using 16-235 for my DVD player going to throw off my grayscale.

Grayscale is really another term for color balance, color tracking, etc., and is a global setting (as far as inputs are concerned) so it would be fine.

As far as colorspace is concerned, and assuming that the projector is properly designed (which it seems to be in this regard) it would make no difference as long as you match colorspace settings on source and projector. So if your Bravo was outputting in the 16-235 space and you were set to a Video GBR mode on your Ruby (which is 16-235) it would track correctly.

Quote:


Also I assume that grayscale aside, I cannot properly calibrate brightness using ColorFacts and the BC for Video sources, since CF presumably will be using 0-255 and my player using 16-235?

If I understand this question correctly, that assumption would be wrong. Remember that in the 16-235 space 16 is black, equivalent to 0-255's 0, and 235 in the 16-235 space is white, just like 255 in the 0-255 space. So as long as your Ruby input colorspace matches the source colorspace your calibrations will stand.

Quote:


I've heard this a lot lately but I do not follow the logic here. Please help me understand this. I understand the point that ideally the pj is calibrated to a spec, and that way it works great as long as the source adheres perfectly to the spec. If I was a pj manufacturer calibrating for D65 out of the factor that is how I would do it.

But to me it makes more sense just to calibrate directly from the DVD player. For instance I use the Bravo D1 to output 720p or 1080i. If I use the Bravo to display the test patterns (instead of CF) and calibrate grayscale off of that, don't I then wind up with a calibration at D65 that is corrected for any errors in my source?

This way if the Bravo has errors I am correcting them to get back to the D65. Now granted this calibration is only applicable for use with my DVD player, but as long as I can use different settings for different sources, logically this seems to me to be the better way to go.

Then how would you calibrate for your HD receiver? See, the whole thing is about standardizing and having a point of reference. It's done this way professionally in both audio and video worlds so you can judge the balance of what you're viewing. In a home environment it's less rigid, of course, but the concept still makes sense there too if you have or plan to have more than one source.

DVD's vary significantly in calibration during production and transfer (mastering), and to a lesser extent, so do HD broadcasts. If your display is calibrated to a standard, you'll find there is much less objectionable deviation between sources. What you'll see more clearly is differences in production style, differences in transfers, obvious faults on the broadcaster's side, etc. If you then choose to fine tune, you can, but you still know where the reference settings are. Without a standard you would be adjusting all the time and never really know how far off or in what manner the source is incorrect.

Does this help any?

--Bill
bblue is offline  
post #94 of 615 Old 01-25-2006, 12:12 PM
AVS Special Member
 
glenned's Avatar
 
Join Date: Aug 2002
Location: Laguna Niguel, CA
Posts: 1,583
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by lovingdvd View Post

But to me it makes more sense just to calibrate directly from the DVD player. . .This way if the Bravo has errors I am correcting them to get back to the D65. Now granted this calibration is only applicable for use with my DVD player, but as long as I can use different settings for different sources, logically this seems to me to be the better way to go.

LovingDVD and BBlue,

This is how calibration has been done traditionally, each source on a different display input and each input calibrated individually. It is ideal to include the source component in the video chain for the reasons that LovingDVD mentioned. With analogue sources this nullifies the variances inherent in analogue designs. Each input on the display should automatically store and recall its own calibration settings whenever it is active. Some displays will store multiple calibrations on a single input depending on whether the signal is Interlaced, Progressive, SD, HD, RGB, or YCbCr.

With digital sources, it looked like this variance was no longer going to be a factor. The first DVI sources put out reference levels (either PC or Video). The Bravo D1 puts out reference level digital signals in the Video signal range (16-235). The D1's output with test patterns from DVE matches the output from my Accupel. (Be aware that the Window Patterns in Avia have some color contamination and do not match exactly, but they are close from 100 IRE down to 30 IRE.)

However, there are newer digital DVD players which do not output reference levels at their factory settings. If they are the only source connected to a particular input on a display, then calibrating through the DVD player nullifies the problem and is the easiest way to make the picture accurate. Alternately, the source can be set to output reference levels by adjusting Contrast and Brightness settings in the DVD player (along with other settings too, depending on the particular DVD player) before calibration. In that case one can either calibrate through the source, or one can use a proxy device which also outputs reference levels, such as an Accupel. The results are the same.

If you want two digital sources to share the same input, then they should both be set to output Video reference signal levels. That way the calibration for one will also be accurate for the other. To do this, you need to start out with a device that you know puts out reference levels (I use an Accupel.) You set the controls on the display to match that source. Then you set the controls on your other sources by sending test patterns through them so that they match the display's settings also. Finally, you want to send gray scale test patterns through both sources to insure that they do indeed match the reference signal levels that were put out by your proxy device.

So it seems to me that you are both right. Two different means to the same end.

Glenn
glenned is offline  
post #95 of 615 Old 01-25-2006, 01:37 PM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Thanks Bill. So in summary if you use ColorFacts to calibrate to D65 over PC levels, and a DVD player matches standard perfectly, then you will still have D65 when later connecting the DVD player (even those the calibration was done in the 0-255 space and the DVD player will be playing back in the 16-235 space)?
lovingdvd is offline  
post #96 of 615 Old 01-25-2006, 02:13 PM
Super Moderator
 
Bob Sorel's Avatar
 
Join Date: Jan 2000
Posts: 8,454
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 63 Post(s)
Liked: 106
Quote:


If you want two digital sources to share the same input, then they should both be set to output Video reference signal levels. That way the calibration for one will also be accurate for the other. To do this, you need to start out with a device that you know puts out reference levels (I use an Accupel.) You set the controls on the display to match that source. Then you set the controls on your other sources by sending test patterns through them so that they match the display's settings also. Finally, you want to send gray scale test patterns through both sources to insure that they do indeed match the reference signal levels that were put out by your proxy device.

Great! Thanks, Glenn! That's the way I figured it should be done, but after reading all the posts in this thread, I was beginning to doubt my sanity. My AccuPel arrived today...

Quote:


However, there are newer digital DVD players which do not output reference levels at their factory settings.

That would seem to be the case with my Panny S-97 and MyHD card...
Bob Sorel is offline  
post #97 of 615 Old 01-25-2006, 06:53 PM
AVS Special Member
 
bblue's Avatar
 
Join Date: Nov 2001
Location: El Cajon, CA
Posts: 1,051
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by JeffY View Post

Well Bill you are confused because you are talking about luma levels yet continue to reference it as colorspace. Just to be clear, 0 - 100 is IRE levels, 16 - 235 is for component luma levels and 16 - 240 is for component chroma levels and 0-255 is for RGB levels.

Ok, Jeff. I'll concede that my use of the term colorspace is not the most accurate. It's more of an encoding/decoding term. But in some circles is used as a general catch-all as I'm using it. I'll try to be more clear in the future.

--Bill
bblue is offline  
post #98 of 615 Old 01-25-2006, 07:01 PM
AVS Special Member
 
bblue's Avatar
 
Join Date: Nov 2001
Location: El Cajon, CA
Posts: 1,051
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by lovingdvd View Post

Thanks Bill. So in summary if you use ColorFacts to calibrate to D65 over PC levels, and a DVD player matches standard perfectly, then you will still have D65 when later connecting the DVD player (even those the calibration was done in the 0-255 space and the DVD player will be playing back in the 16-235 space)?

Correct.

Except that you're actually talking about two different things. D65 is a color temperature reference, the result of balancing colors, and has (for this conversation) nothing to do with levels. With the levels you're making sure that black output (0% or 0 IRE) is really at black, and white output (100% or 100 IRE) is really at white. That is independent of temperature.

--Bill
bblue is offline  
post #99 of 615 Old 01-25-2006, 08:06 PM
AVS Special Member
 
ddingle's Avatar
 
Join Date: May 2001
Location: Mahtomedi,Minnesota
Posts: 1,098
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by ddingle View Post

I contacted Milori. They do not sell or supply the eye one with cosine correction or whatever. They recommended contacting Gretag. Milori indicated they will however support it? I left a message with Gretag earlier today. Hopefully some answers will be forthcoming. Pricing etc.

Gretag called back. No help really. I either did not explain very well or he did not know much about the product line.Or both.
Although as mentioned earlier in the thread UMR has a website Accucal that sells a software to use with the eye one pro. Starting at $500. The eye one pro is extra. There may be a trade in on your existing unit,but the $750 price is roughly the price of the sensor. Since I own the Colorfacts already I am uneasy about dumping more money into software and sensors. Colorfacts mentioned support for the eye one pro. I guess I will go back to them and see what they can offer. The whole process seems about as clear as mud.
ddingle is offline  
post #100 of 615 Old 01-25-2006, 09:35 PM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Quote:
Originally Posted by ddingle View Post

Gretag called back. No help really. I either did not explain very well or he did not know much about the product line.Or both.
Although as mentioned earlier in the thread UMR has a website Accucal that sells a software to use with the eye one pro. Starting at $500. The eye one pro is extra. There may be a trade in on your existing unit,but the $750 price is roughly the price of the sensor. Since I own the Colorfacts already I am uneasy about dumping more money into software and sensors. Colorfacts mentioned support for the eye one pro. I guess I will go back to them and see what they can offer. The whole process seems about as clear as mud.

Ditto that. I spent a few hours today wading around in the same mud. No one seems to know anything about what we are talking about. Gretag didn't know what I was talking about. Diffuser? What diffuser?

They say the new EyeOne "rev b" comes with an "ambient light hood", or something like that. I assume that must be the piece we need to diffuse the light for reading directly from the sensor - mainly because they said it has to be calibrated in conjunction with each EyeOne sensor which someone here earlier said would likely be the case.

Anyway I agree with you completely in that I am very hesitant to dump about $700 into a Gretag product that's not working so great for me in the first place. It may be worth it if it solves both my issues (poor readings below 30 IRE and inability to measure primaries accurately) but who knows if that'll be the case...
lovingdvd is offline  
post #101 of 615 Old 01-27-2006, 06:47 AM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Quote:
Originally Posted by Bob Sorel View Post

The EyeOne Pro is the sensor itself - "Beamer" is a software/accessory package that is bundled with the EyeOne Pro sensor. I also bought the "Beamer" package, as it includes a tripod mount and more importantly, a diffuser for taking readings directly from the light source. You can not buy the diffuser separately.

Bob - Its my understanding that the EyeOne Pro cannot be used on the tripod mount when the ambient light hood (diffuser) is attached to the unit. Is this correct?

All - does anyone know whether ColorFacts 6.0 supports the EyeOne Pro in the ambient light mode so you can do calibrations with the diffuser? See http://www.avsforum.com/avs-vb/showthread.php?t=636633 .
lovingdvd is offline  
post #102 of 615 Old 01-28-2006, 05:40 PM
AVS Special Member
 
glenned's Avatar
 
Join Date: Aug 2002
Location: Laguna Niguel, CA
Posts: 1,583
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by lovingdvd View Post

Bob - Its my understanding that the EyeOne Pro cannot be used on the tripod mount when the ambient light hood (diffuser) is attached to the unit. Is this correct?

All - does anyone know whether ColorFacts 6.0 supports the EyeOne Pro in the ambient light mode so you can do calibrations with the diffuser? See http://www.avsforum.com/avs-vb/showthread.php?t=636633 .

The cosign corrector obstructs the built in mount. I made a cup from ABS plastic pipe which holds the eye-one. There is a nut cemented to the cup for attaching to a standard tripod head.

Glenn
glenned is offline  
post #103 of 615 Old 01-30-2006, 08:13 AM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Quote:
Originally Posted by glenned View Post

The cosign corrector obstructs the built in mount. I made a cup from ABS plastic pipe which holds the eye-one. There is a nut cemented to the cup for attaching to a standard tripod head.

Glenn

Sounds like a clever design Glenn. Can you post a picture?
lovingdvd is offline  
post #104 of 615 Old 01-30-2006, 08:18 AM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
I recently became aware that the Lumagen has the ability to calibrate grayscale with R,G,B gain/offset values that can be set independantly in 10 IRE steps.

Am I correct to assume this would be very helpful in getting a perfect D65 with a dE of say 1 or less in the Ruby's auto iris mode?

If I'm understand this correctly, it would be very helpful because you could then easily tweak the high end which gets affected the most with color shift due to the dynamic iris, without affecting the mid to low end. Much less of a balance act required. What do you think?

At any rate I wouldn't see this is mandatory, as I managed to get D65 within dE of 3 or less from 30-100 (couldn't measure below 30 accurately) just by using the Ruby's controls. But nonetheless this sounded like some great flexibility by using the Lumagen...
lovingdvd is offline  
post #105 of 615 Old 01-30-2006, 08:21 AM
AVS Club Gold
 
Alan Gouger's Avatar
 
Join Date: Jan 1999
Location: Florida
Posts: 18,726
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 16
Quote:
Originally Posted by lovingdvd View Post

I recently became aware that the Lumagen has the ability to calibrate grayscale with R,G,B gain/offset values that can be set independantly in 10 IRE steps.

Am I correct to assume this would be very helpful in getting a perfect D65 with a dE of say 1 or less in the Ruby's auto iris mode?

If I'm understand this correctly, it would be very helpful because you could then easily tweak the high end which gets affected the most with color shift due to the dynamic iris, without affecting the mid to low end. Much less of a balance act required. What do you think?

At any rate I wouldn't see this is mandatory, as I managed to get D65 within dE of 3 or less from 30-100 (couldn't measure below 30 accurately) just by using the Ruby's controls. But nonetheless this sounded like some great flexibility by using the Lumagen...

It also has four memories that would work out well for the Ruby. You can also add your own gamma curve instead of using the Sony software.
Alan Gouger is offline  
post #106 of 615 Old 01-31-2006, 06:26 PM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Quote:
Originally Posted by Bytehoven View Post

I have some interesting news regarding gamma shading.

I just got back my HS-51 from the Sony Laredo site, which had been sent in for a grayscale/white balance uniformity issue. My projector had developed a severe red/blue left/right bias.

They replaced the prism block, which includes the the LCD panels. They recalibrated the LOW, MID & HIIGH color temp values for the new panels. Then they also said they applied special gamma shading.

I called to get more specifics on the repair and they told me a little bit more about the gamma shading utility. They said the utility works on 7 video levels and on up to 256 points on the screen. The utility apparently has versions for the newer HS-51A/60 as well as the VW100.

I'm checking with some other Sony contacts to see if this gamma shading utility is available to third party calibrators, or just as a factory/service tool. I assume it's only the latter, but one can hope.

Incidently, The Sony Laredo site was able to bring my HS-51 back to spec.


Any feel as to what benefits a gamma shading utility might bring to the table? What would something like this be used for? To temper the brightness in Ruby's corners perhaps?
lovingdvd is offline  
post #107 of 615 Old 01-31-2006, 06:41 PM
AVS Special Member
 
sfogg's Avatar
 
Join Date: Jan 2001
Location: Ma, USA
Posts: 5,614
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
"Any feel as to what benefits a gamma shading utility might bring to the table? What would something like this be used for? "

Ability to help correct for color uniformity issues. Put up a white 100 IRE screen... is it uniformly white? Work your way down 10 IRE at a time and see if you get points with more green/blue/red on certain parts of the screen. If so that is a shading problem. A shading utility can possibly even that out depending upon how much shading control is built into the projector.

Shawn
sfogg is offline  
post #108 of 615 Old 02-01-2006, 03:46 PM
Advanced Member
 
ericlhyman's Avatar
 
Join Date: May 2001
Location: Silver Spring, MD USA
Posts: 840
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Has anyone used the gamma adjustment disk that came with the Ruby? Can someone provide instructions for this?

ericlhyman
ericlhyman is offline  
post #109 of 615 Old 02-01-2006, 04:50 PM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Quote:
Originally Posted by ericlhyman View Post

Has anyone used the gamma adjustment disk that came with the Ruby? Can someone provide instructions for this?

I posted a thread several weeks ago to pose the same question but got little or no response. I haven't really pursed it since. I did install the software on my PC and interestingly the docs seem to explain more than I thought they would.
lovingdvd is offline  
post #110 of 615 Old 02-01-2006, 06:40 PM
Advanced Member
 
ghibliss's Avatar
 
Join Date: Aug 1999
Location: New York, NY
Posts: 740
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
GregR,

The CA-6X analyzer probe which we sell (also used in the OpticOne which we made for Ovation) provides a highly accurate calibration on Lcos displays using either UHP or Xenon lamps. William Phelps can vouch for the accuracy of this device as he has been using this probe for his JVC display optimizations and has found it to be within .002 for xy accuracy relative to his LightSpex spectroradiometer. I doubt that anyone will find issues with this degree of error in a display calibration!

Cliff Plavin
Progressive Labs
ghibliss is offline  
post #111 of 615 Old 02-01-2006, 07:24 PM
AVS Special Member
 
gregr's Avatar
 
Join Date: Nov 2001
Posts: 2,127
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by ghibliss View Post

GregR,

The CA-6X analyzer probe which we sell (also used in the OpticOne which we made for Ovation) provides a highly accurate calibration on Lcos displays using either UHP or Xenon lamps. William Phelps can vouch for the accuracy of this device as he has been using this probe for his JVC display optimizations and has found it to be within .002 for xy accuracy relative to his LightSpex spectroradiometer. I doubt that anyone will find issues with this degree of error in a display calibration!

Cliff Plavin
Progressive Labs

That's good to know, Cliff. I do appreciate that information. But just so you know, I never said it would have a problem. I just mentioned that Xenon lamps produce a different spectral distribution, so tri-stimulus color analyzers may have larger errors with Xenon lamps than they exhibit with the more common UHP lamps.

Greg Rogers
Video Engineer/Product Designer

gregr is offline  
post #112 of 615 Old 02-03-2006, 09:01 AM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
I recently got a SpyderTV (STV) meter because of difficulties getting accurate grayscale readings below 30 IRE with my EyeOne and ColorFacts. Also I was hoping that the STV would allow me to finally read primaries accurately - it did not, but that's a story for another day.

The first thing I noticed was that the STV was really struggling to take readings off the screen at lower light levels. This was contrary to my understanding of the meter. According to DataColor the STV for front pjs should be used emissively (read off the screen) and NOT pointed at the pj. It was not working out well for me like this at all.

Other AVS members told me they point their STV directly at the pj, so I decided to give it a try. I was surprised (based on DataColor's advice) to see the STV did extremely well reading directly from the pj. Much better than the EyeOne off the screen, and certainly much better than its own readings off the screen.

With this squared away, I moved on to comparing high IRE grayscale readings with the EyeOne to STV. There was about a 500K difference or so between the two meters. Since it is my understanding that the EyeOne is a higher end device, and that some tri-stimulus meters may potentially not be as accurate on SXRD, I decided to trust the EyeOne as being more accurate.

So with that conclusion, I then used CF 6's Train Meter function to train the STV to the EyeOne. I did this by taking measurements from the EyeOne off the screen, and then from STV pointed directly at the pj. For further accuracy, I also increased the exposure time in CF options to the maximum (8000 ms I believe).

Next I proceeded to recalibrate the Ruby. Last calibration was at 40 hours, and now I am at 80 hours. I was surprised to see just how much things have moved in only 40 hours as the final results of the calibration have much different numbers. This is not a result of using the new meter - I even checked it with the EyeOne.

One thing that was great is that now with the STV I not only could read 20 IRE accurately, but also 10 IRE! Although, in auto-iris mode, I had to move the STV closer to the pj (within about 4 feet) in order to get accurate readings at 10 IRE.

I was very surprised to see just how off my grayscale was from the last calibration at 10 IRE. It was at 7600 with a dE of about 17 I think. 20 IRE on the other hand was near D65 dE 2.

As you may recall, in my previous posts I said I had 30-100 IRE at D65 because I could not measure below 30. Well now that I can I see how off it was at 10 IRE. With a 10 IRE window up this was also obvious by eye, but I hadn't really looked at a 10 IRE window previously because I knew I couldn't read it with the EyeOne.

I've been calibrating for several years now, but this was the first time I could read 10 IRE. What was really surprising to me was just how sensitive the grayscale is down at 10 IRE. For instance there was one exact spot for R,G,B bias that put me near D65 with a dE of 1. Any single click adjustment of any color immediately jumped me to dE of 7+. This was not the STV causing this. Rather you could even see it by eye start to turn color.

Now that I could accurately read 10 IRE, I was determined to have it dialed in correctly. It was very challenging to have 10 IRE dialed in correct AND 20-50 IRE also dialed in correctly.

In fact, it reached a point where it was impossible to dial in D65 0-3 dE from 10-50 IRE. The closest I could get it was dE of 7 at 10 IRE with dE of 1 from 20-50 IRE. It was either that, or have a dE of 1 at 10 IRE with a dE of 4 from 20-50 IRE. I opted for the former. I figured I'd rather have a perfect grayscale from 20-50 with it a little off a 10 IRE, then to have it perfect at 10 IRE and be a little off from 20-50. Question - which would you guys have opted for given those choices? I know a dE of 7 is fairly significant, but at 10 IRE is it noticeable enough to worry about given the alternative is dE of 4 from 20-50 IRE?

It would certainly have been great to be using a processor like the Lumagen which allows you to fine tune grayscale calibration in 10 IRE steps. Using this technique I think I could be at D65 from 10-100, with a dE of 0 from 10-70 and dE of 0-3 from 80-100 IRE (auto iris plays a factor here).

After this experience I'm going to create a pattern to help make any errors at 10 IRE more obvious. I'll create horizontal bands at 25% screen height each to create 4 steps at 0, 10, 20 and 30 IRE. I know there are some pre-existings ramps for this type of testing, but to me those ramp patterns do not make errors on the very low end obvious enough.

Also I found it even more challenging than before to dial in the auto iris mode. I'm not sure, but I think using Darin's factory iris setting tweaks may have made this more difficult than the last calibration when I was using just the factory iris settings at that time. Eventually I got it down to 20-100 IRE with a dE of 0-3, and dE of 7 at 10 IRE (which had nothing to do with the iris). The ironic thing is that the final gains/bais settings between the iris auto and off modes was very minimla (see results at end of this post) compared to the last time. So the numbers were much closer, but it was tougher to get within dE of 3 despite this as slight changes caused too much error one way or the other/

Question: Is it typical in digital displays for things to be great from say 20-100 IRE but things then fall apart at 10 IRE and lower unless you are specifically careful about this and purposely work on it? In other words, do things tend to track nicely from 20-100 but not so much so at 10 IRE and below? Since this was my first experience reading below 20 IRE I have no prior observations to base this on.

BTW, for those of you that watch with this pj in iris OFF mode - here's a quick test. Put up a 10 IRE window (or even a 20) and switch back and forth between iris auto and iris off. The difference in black level is HUGE (much better blacks in auto mode of course). After seeing this you may find yourself switching to the auto iris for your serious viewing.

Here are my final numbers. Note that I did not calibrate the iris ON position this time since I never watch in this mode.

Contrast 80, Brightness 50, Lamp at 85 hours

____IRIS OFF | AUTO

Gain R: _98 | _99
Gain G: 111 | 110
Gain B: 126 | 128

Bias R: 129 | 130
Bias G: 126 | 126
Bias B: 130 | 130


Questions and comments always welcome.
lovingdvd is offline  
post #113 of 615 Old 02-03-2006, 09:18 AM
umr
AVS Club Gold
 
umr's Avatar
 
Join Date: Oct 2002
Location: USA
Posts: 10,134
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 29
Quote:
Originally Posted by lovingdvd View Post

....Question - which would you guys have opted for given those choices? I know a dE of 7 is fairly significant, but at 10 IRE is it noticeable enough to worry about given the alternative is dE of 4 from 20-50 IRE?

....Question: Is it typical in digital displays for things to be great from say 20-100 IRE but things then fall apart at 10 IRE and lower unless you are specifically careful about this and purposely work on it? In other words, do things tend to track nicely from 20-100 but not so much so at 10 IRE and below? Since this was my first experience reading below 20 IRE I have no prior observations to base this on.....

I would answer your first question by saying that calibrating by dE alone is a huge mistake. You must be sure that the color errors are in a direction that is tolerable. A green error is much worse than a blue error. People with too much green in their skin just look wrong. Without looking at some images it is impossible to say what is the correct way to calibrate your projector. This is not to say that I am against using dE. It is a great tool, but only one of several that need to be considered in the calibration process.

With respect to your second question it is common with most displays for things to be reasonable from 20 to 100 IRE and fall apart at 10 IRE and less. As I said before this must be considered in the calibration whether it can be measured with an instrument or not. Once the brighter steps are properly calibrated color shifts are easier to see in the lower levels. Noise will also frequently increase at these lower levels and should be minimized during the calibration if possible.
umr is offline  
post #114 of 615 Old 02-03-2006, 09:44 AM
Member
 
thomasclaus's Avatar
 
Join Date: Sep 2002
Location: Germany
Posts: 91
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Hi,

Quote:


I would answer your first question by saying that calibrating by dE alone is a huge mistake. You must be sure that the color errors are in a direction that is tolerable.

Not only that, but CCT does not provide a definite information about the R,G,B portions, too. One may achieve a certain CCT value by various combinations of R,G,B. Therefore I almost completely ignore the CCT; my reference is a balance of R,G,B, only.

I would not be too surprised that the RGB-levels are not dead-on for the complete range of brightness from 0 to 100 IRE - especially if you tweaked Iris, Gamma, Bias and Gain already. Fine tuning for dedicated IRE steps may be done with the Image Director software - what a pity that this software is so bugged. I wished someone will be able to do a read-out of the save-files so one could "construct" its own Gamma curve values with Excel or a proprietary little piece of software without being bothered with those ugly limitations of the ID 2.30 Software.

If I only had more time for all this...

Thomas

The saw is family
thomasclaus is offline  
post #115 of 615 Old 02-03-2006, 09:48 AM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Quote:
Originally Posted by umr View Post

I would answer your first question by saying that calibrating by dE alone is a huge mistake. You must be sure that the color errors are in a direction that is tolerable. A green error is much worse than a blue error. People with too much green in their skin just look wrong. Without looking at some images it is impossible to say what is the correct way to calibrate your projector. This is not to say that I am against using dE. It is a great tool, but only one of several that need to be considered in the calibration process.

Thanks Jeff. That's a great point I hadn't considered. In this case, IIRC, I was fighting back and forth over one click of blue bias. If I raised this from 130 to 131, then I was at dE 1 at 10 IRE but at dE 4 from 20-50 IRE because 20-50 now had a bit too much blue to stay within my dE 3 target.

So with your answer in mind, it sounds like going to dE of 4 from 20-50 because blue is a tad to high, would be tolerable since you said its not as noticeable.

Now I'm trying to recall what things looked like at 10 IRE with dE of 7..So I know for sure it was a bit short on blue (because one click increase in blue got me to dE of 1). I can't recall but it might have just been red and green at about the same level say 101-102% with blue at like 94%. I think the CTT was about 6350K.

So knowing this does that help you make an determination in which way would have been better.

If it makes any difference, I did go back and forth with a 10 IRE pattern with the 1 click of blue back and forth to visually gauge the difference this one click made to the gray at 10 IRE. I remember it looking a tad warm at 130 vs 131 on the blue bias. It was enough to make a difference if you really stared at it and went back and forth serveral times. If anything it was a tad warm at dE 7.

Quote:


With respect to your second question it is common with most displays for things to be reasonable from 20 to 100 IRE and fall apart at 10 IRE and less. As I said before this must be considered in the calibration whether it can be measured with an instrument or not. Once the brighter steps are properly calibrated color shifts are easier to see in the lower levels. Noise will also frequently increase at these lower levels and should be minimized during the calibration if possible.

Is it also common so the slightest adjustment to make such a significant impact at 10 IRE? In other words, at say 20 or 30 IRE, one click in blue generally only shifts the CTT a little. But it seems like at 10 IRE on click of any color in any direction has a very significant impact on CCT. Is that normal or do you think this is more about the accuracy of the meter reading at 10 IRE? It didn't seem like an accuracy issue though, because you really can see the color shifting a bit by eye as you make a single click. Now that I've seen this I'm very interested in adding 10 IRE step calibration control from the Lumagen.
lovingdvd is offline  
post #116 of 615 Old 02-03-2006, 09:53 AM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Quote:
Originally Posted by thomasclaus View Post

Hi,



Not only that, but CCT does not provide a definite information about the R,G,B portions, too. One may achieve a certain CCT value by various combinations of R,G,B. Therefore I almost completely ignore the CCT; my reference is a balance of R,G,B, only.

Hi Thomas. I thought that by very definition, if you are at D65 then R,G,B are all at the same level (100%).

When I calibrate I mostly just watch dE and try to target 0, but will accept up to 3. If it becomes impossible to stay below dE of 3 at all levels, I'll now follow Jeff's advice to see how acceptable a dE of 4+ is based on what color its shifted toward.

But the bottom line is that it is correct just to follow this method right? My understanding is that dE of 3 or less is generally not perceivable, so as long as that's there any there's any slight color balance error (say 102%, 101%, 100%) it doesn't really matter as long as its at dE of 0. At least that what it sounds like if I'm following all these principles.
lovingdvd is offline  
post #117 of 615 Old 02-03-2006, 09:59 AM
Super Moderator
 
Bob Sorel's Avatar
 
Join Date: Jan 2000
Posts: 8,454
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 63 Post(s)
Liked: 106
Hi Ric and Jeff,

I would like to calibrate my Ruby using my newly purchased AccuPel, but I am really unsure of the procedure. I have read posts from Greg talking about switching back and fourth between windowed and full field patterns when calibrating the auto iris mode, but I have no idea how to use that information in the calibration process. I assume that because windowed patterns change the APL, it affects how open or closed the iris is, and that in turn would affect the gamma curve and grayscale, but it seems to me that you can only get one set right (windowed or full field) and the other set will be wrong. So how do you use both sets to achieve proper calibration? Besides, I don't think that the Accupel has a complete set of full field gray patterns.
Bob Sorel is offline  
post #118 of 615 Old 02-03-2006, 10:01 AM
umr
AVS Club Gold
 
umr's Avatar
 
Join Date: Oct 2002
Location: USA
Posts: 10,134
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 29
Quote:
Originally Posted by lovingdvd View Post

....So knowing this does that help you make an determination in which way would have been better

....Is it also common so the slightest adjustment to make such a significant impact at 10 IRE?

...Now that I've seen this I'm very interested in adding 10 IRE step calibration control from the Lumagen.

I cannot make a determination for you on how to set your gray scale. It requires me to see it. Measurements alone are insufficient. There could be other problems that would impact my choices. I would suggest that blue errors should be favored over red and green. I would then favor a red error over green if I could not go with blue.

It is very common that small changes will strongly effect the color performance from 0-10 IRE. It should be obvious that this would be the case. You are moving a bias which by its very nature is going to effect the results near zero the greatest percentage.

I would not assume that the Lumagen will improve this situation much. You may be better off searching for RGB gamma curves within the Ruby. They are likely to have more resolution.
umr is offline  
post #119 of 615 Old 02-03-2006, 02:07 PM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Quote:
Originally Posted by umr View Post

...
I would not assume that the Lumagen will improve this situation much. You may be better off searching for RGB gamma curves within the Ruby. They are likely to have more resolution.

Thanks for all the input Jeff. Regarding the Lumagen - why would you think it may not improve things? As I understand it, it provides the ability to set R/G/B gain and bias independantly at 10 IRE steps. So for instance, if I have the Lumagen, I could simply have gone with the dE of 1 from 20-50 using the Ruby controls, and then used the lumagen controls specifically at 10 IRE to add a little blue and get me much closer to dE of 0 at 10 IRE without affecting really impacting 20-50, right? Or am I missing something?
lovingdvd is offline  
post #120 of 615 Old 02-03-2006, 02:13 PM - Thread Starter
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,818
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 23
Quote:
Originally Posted by Bob Sorel View Post

Hi Ric and Jeff,

I would like to calibrate my Ruby using my newly purchased AccuPel, but I am really unsure of the procedure. I have read posts from Greg talking about switching back and fourth between windowed and full field patterns when calibrating the auto iris mode, but I have no idea how to use that information in the calibration process. I assume that because windowed patterns change the APL, it affects how open or closed the iris is, and that in turn would affect the gamma curve and grayscale, but it seems to me that you can only get one set right (windowed or full field) and the other set will be wrong. So how do you use both sets to achieve proper calibration? Besides, I don't think that the Accupel has a complete set of full field gray patterns.

Bob - See my post here: http://www.avsforum.com/avs-vb/showt...&&#post6961119 as I explain in some pretty good detail the approach to calibrating the dynamic iris and provide some good tips and tricks.

I know it sounds crazy, but basically you really can go back and forth between full fields and varying sized windows and have 20-100 within 0-3 dE, maybe with a spot at dE 4 and 10 IRE at say dE 1-7 depending on how you go about it. You will reach a point where you have to sacrifice a little - such as deciding whether you want a dE of 4 on a window vs full field - can't always have it both ways but like I said you can get it very very close. Any questions just post them - I'll be happy to help.

What sensor and software will you be using to calibrate?
lovingdvd is offline  
Reply Digital Hi-End Projectors - $3,000+ USD MSRP

Tags
Sony Vpl Vw100 Lcd Projector
Gear in this thread

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off